Case file · NAS & RAID · SDR-2025-0914
The Cloud Disk Came Out and Still Won't Read.
When the My Cloud stopped answering, the owner worked through the sensible steps in order: took the disk out, wired it straight into a PC, and armed Windows with a program that lets Windows read linux drives
. It changed nothing — still can't get at the files
. Nearly all of it had a copy elsewhere; the months that did not were the whole problem.
Seeing the same thing yourself?
0800 6890668
The translation.
Linux formatting was the obvious culprit, and the owner's own test ruled it out — genuinely useful legwork. That leaves something further down: either the disk is not returning reads dependably, or the structures describing the volume are damaged. A My Cloud adds a wrinkle of its own, because WD stacks its own arrangement on top of ordinary Linux storage, so a tool that happily opens a plain ext4 partition can still find nothing there it knows how to open.
Kit used on this job.
What happens in a case →| Platform | Its role in this case | Why we use it |
|---|---|---|
| DeepSpar Disk Imager 4 | Imaged the disk with timeout handling no desktop program could manage | Head-map-driven imaging, with per-head timeouts, resets and power-cycle control |
| UFS Explorer RAID Recovery | Unpicked how the NAS lays its volume out, below filesystem level | Understands how NAS volume managers really work, instead of a flat array |
| R-Studio Technician | Compared the recovered tree with the owner's own account of what belonged there | Wide filesystem support, with RAID rebuilds you can trust |
In the lab.
Verify the ground his own testing had already covered
On the bench the disk read unevenly — some requests answered, others hanging, with no pattern the desktop tools could make sense of. Behaviour like that will defeat any program running on an ordinary operating system, because such software retries a handful of times and then walks away.
Take a hardware-level copy, slowly
With timeouts held short and stalls answered by a reset rather than a surrender, the disk gave up nearly all of itself. Weak areas were logged instead of wrestled with, then returned to at the end once everything straightforward was already secure.
Peel back the NAS layers, not the filesystem alone
Working from the image, the volume was taken apart properly rather than treated as an ordinary partition — first the arrangement WD's firmware lays over standard Linux storage, then the filesystem sitting inside that. The shares came up in their right shape, and the months that had never been copied were found.
The result.
The uncopied months came back and rejoined the material that had been duplicated. His orderly elimination had genuinely saved diagnostic time; the patient imaging and the layer-by-layer work covered the rest.
Similar cases on the index.
Also from NAS & RAID.
Does this sound like your own drive?
The rule holds as in every case above: switch it off, and let a free diagnosis come before any decision.